Files
orion/app/modules/prospecting/routes/api/admin_enrichment.py
Samir Boulahtit 22ae63b414
Some checks failed
CI / validate (push) Has been cancelled
CI / ruff (push) Successful in 10s
CI / dependency-scanning (push) Has been cancelled
CI / docs (push) Has been cancelled
CI / deploy (push) Has been cancelled
CI / pytest (push) Has started running
refactor(prospecting): migrate SVC-006 transaction control to endpoint level
Move db.commit() from services to API endpoints and Celery tasks.
Services now use db.flush() only; endpoints own the transaction.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-02-28 16:40:09 +01:00

187 lines
6.5 KiB
Python

# app/modules/prospecting/routes/api/admin_enrichment.py
"""
Admin API routes for enrichment/scanning pipeline.
"""
import logging
from fastapi import APIRouter, Depends, Path, Query
from sqlalchemy.orm import Session
from app.api.deps import get_current_admin_api
from app.core.database import get_db
from app.modules.prospecting.schemas.enrichment import (
ContactScrapeResponse,
FullEnrichmentResponse,
HttpCheckBatchItem,
HttpCheckBatchResponse,
HttpCheckResult,
ScanBatchResponse,
ScanSingleResponse,
ScoreComputeBatchResponse,
)
from app.modules.prospecting.services.enrichment_service import enrichment_service
from app.modules.prospecting.services.prospect_service import prospect_service
from app.modules.prospecting.services.scoring_service import scoring_service
from app.modules.tenancy.schemas.auth import UserContext
router = APIRouter(prefix="/enrichment")
logger = logging.getLogger(__name__)
@router.post("/http-check/{prospect_id}", response_model=HttpCheckResult)
def http_check_single(
prospect_id: int = Path(...),
db: Session = Depends(get_db),
current_admin: UserContext = Depends(get_current_admin_api),
):
"""Run HTTP connectivity check for a single prospect."""
prospect = prospect_service.get_by_id(db, prospect_id)
result = enrichment_service.check_http(db, prospect)
db.commit()
return HttpCheckResult(**result)
@router.post("/http-check/batch", response_model=HttpCheckBatchResponse)
def http_check_batch(
limit: int = Query(100, ge=1, le=500),
db: Session = Depends(get_db),
current_admin: UserContext = Depends(get_current_admin_api),
):
"""Run HTTP check for pending prospects."""
prospects = prospect_service.get_pending_http_check(db, limit=limit)
results = []
for prospect in prospects:
result = enrichment_service.check_http(db, prospect)
results.append(HttpCheckBatchItem(domain=prospect.domain_name, **result))
db.commit()
return HttpCheckBatchResponse(processed=len(results), results=results)
@router.post("/tech-scan/{prospect_id}", response_model=ScanSingleResponse)
def tech_scan_single(
prospect_id: int = Path(...),
db: Session = Depends(get_db),
current_admin: UserContext = Depends(get_current_admin_api),
):
"""Run technology scan for a single prospect."""
prospect = prospect_service.get_by_id(db, prospect_id)
profile = enrichment_service.scan_tech_stack(db, prospect)
db.commit()
return ScanSingleResponse(domain=prospect.domain_name, profile=profile is not None)
@router.post("/tech-scan/batch", response_model=ScanBatchResponse)
def tech_scan_batch(
limit: int = Query(100, ge=1, le=500),
db: Session = Depends(get_db),
current_admin: UserContext = Depends(get_current_admin_api),
):
"""Run tech scan for pending prospects."""
prospects = prospect_service.get_pending_tech_scan(db, limit=limit)
count = 0
for prospect in prospects:
result = enrichment_service.scan_tech_stack(db, prospect)
if result:
count += 1
db.commit()
return ScanBatchResponse(processed=len(prospects), successful=count)
@router.post("/performance/{prospect_id}", response_model=ScanSingleResponse)
def performance_scan_single(
prospect_id: int = Path(...),
db: Session = Depends(get_db),
current_admin: UserContext = Depends(get_current_admin_api),
):
"""Run PageSpeed audit for a single prospect."""
prospect = prospect_service.get_by_id(db, prospect_id)
profile = enrichment_service.scan_performance(db, prospect)
db.commit()
return ScanSingleResponse(domain=prospect.domain_name, profile=profile is not None)
@router.post("/performance/batch", response_model=ScanBatchResponse)
def performance_scan_batch(
limit: int = Query(50, ge=1, le=200),
db: Session = Depends(get_db),
current_admin: UserContext = Depends(get_current_admin_api),
):
"""Run performance scan for pending prospects."""
prospects = prospect_service.get_pending_performance_scan(db, limit=limit)
count = 0
for prospect in prospects:
result = enrichment_service.scan_performance(db, prospect)
if result:
count += 1
db.commit()
return ScanBatchResponse(processed=len(prospects), successful=count)
@router.post("/contacts/{prospect_id}", response_model=ContactScrapeResponse)
def scrape_contacts_single(
prospect_id: int = Path(...),
db: Session = Depends(get_db),
current_admin: UserContext = Depends(get_current_admin_api),
):
"""Scrape contacts for a single prospect."""
prospect = prospect_service.get_by_id(db, prospect_id)
contacts = enrichment_service.scrape_contacts(db, prospect)
db.commit()
return ContactScrapeResponse(domain=prospect.domain_name, contacts_found=len(contacts))
@router.post("/full/{prospect_id}", response_model=FullEnrichmentResponse)
def full_enrichment(
prospect_id: int = Path(...),
db: Session = Depends(get_db),
current_admin: UserContext = Depends(get_current_admin_api),
):
"""Run full enrichment pipeline for a single prospect."""
prospect = prospect_service.get_by_id(db, prospect_id)
# Step 1: HTTP check
enrichment_service.check_http(db, prospect)
# Step 2: Tech scan (if has website)
tech_profile = None
if prospect.has_website:
tech_profile = enrichment_service.scan_tech_stack(db, prospect)
# Step 3: Performance scan (if has website)
perf_profile = None
if prospect.has_website:
perf_profile = enrichment_service.scan_performance(db, prospect)
# Step 4: Contact scrape (if has website)
contacts = []
if prospect.has_website:
contacts = enrichment_service.scrape_contacts(db, prospect)
# Step 5: Compute score
db.refresh(prospect)
score = scoring_service.compute_score(db, prospect)
db.commit()
return FullEnrichmentResponse(
domain=prospect.domain_name,
has_website=prospect.has_website,
tech_scanned=tech_profile is not None,
perf_scanned=perf_profile is not None,
contacts_found=len(contacts),
score=score.score,
lead_tier=score.lead_tier,
)
@router.post("/score-compute/batch", response_model=ScoreComputeBatchResponse)
def compute_scores_batch(
limit: int = Query(500, ge=1, le=5000),
db: Session = Depends(get_db),
current_admin: UserContext = Depends(get_current_admin_api),
):
"""Compute or recompute scores for all prospects."""
count = scoring_service.compute_all(db, limit=limit)
db.commit()
return ScoreComputeBatchResponse(scored=count)